STUDIES IN JAPANESE by Harumi Befu

1. Introduction In the last several decades, extensive research has resulted in considerable understanding of such areas of Japanese kinship as the nature of the stem (ie), the structure of the dozoku, usages of , changes in family forms and their associated values, and expressions of quasi-familistic ideology in nonkinship settings, as in ritual kinship and industrial organization. Because of limitations of space, I shall only high- light a selective sampling of problems and issues in Japanese kinship, providing minimal references.

2. Dozoku One of the controversies over the nature of the dozoku concerns the question of descent. A number of scholars (Befu 1963; Gamo 1962; Kitano 1966; Okawa 1939) have regarded the dozoku as basically patrilineal. Others (Ariga 1947; Nakane 1967; Nakano 1957) have partially rejected this position inlight of the fact that the dozoku recruits nonkinsmen as a matter of course. They reason that if membership can be recruited through other means than birth, the group cannot be a descent group. If, on the other hand, as Brown (1966) has recently argued, nonkin members are not entitled to all the privileges of "true" members, then there is still room to claim the dozoku as a kinship group and to consider the nature of descent. Brown claims that the descent rule of the dozoku "in ideology" is cognatic rather than patri- lineal. Brown recognizes that in practice-at the "phenomenal levelH-the composition of the dozoku tends statistically to contain more patrikin than other types of kinsmen. His formulation challenges the contention of patri- lineality of earlier workers, whose analyses are perhaps one-sided, based either on the phenomenal order of dozoku groups or on their ideology. Another controversy has to do with the content or substance of the dozoku. Those disavowing the relevance of kinship to the dozoku consider it basically a localized corporate group of with little raison #&re other than its economic and political functions. For this reason many Japanese social scientists restrict use of the term dozoku to groups clearly performing these functions. It then follows that industrialization has con- 114 RICE UNIVERSITY STUDIES tributed to the decline of the dozoku weakening it functionally (Nakane 1967: 172). To Nakano, however, the dozoku is more than a group with certain economic and political functions, which he regards as simply means for perpetuating the dozoku as a corporation (personal communication). What the essential nature of this corporation is, Nakano does not make clear, Brown agrees with Nakano in disclaiming as essential these functional aspects of the dozoku; to him, the dozoku has a substance all its own in the ideological order, providing a sense of mutual solidarity among members. Thus Brown (1968: 135) maintains that industrialization has not necessarily brought about the decline of the dozoku. 3. Domestic Cycle Studies of the family cycle in Japan go back as far as the 1940's (Suzuki 1942). One of the most notable contributions in recent years is Koyama's analysis (1959) of family registries (Shumon nimbetsucho) of a community in Kai Province for 59 years ending in 1861. The methodological importance of this work lies in the fact that it is, I believe, the only study in which con- tinuous, longitudinal records of a group of are subjected to analysis for the purpose of outlining the developmental cycle. (Arthur P. Wolf is now attempting a similar study using family registries from Taiwan.) Pre- vious studies had produced a structural typology of families examined at one point of time. Changes through time are then inferred; type I is assumed to change into type 11, which then is assumed to transform into type 111, and so on. Koyama's longitudinal analysis, however, shows no single domestic cycle in the community. He classifies families into seven types, but for each type he found that change could be to many other types, thus creating a bewildering number of different sequential types of cycles. His findings caution against conveniently arranging synchronic data into a sequential classification of family types and calling them the domestic cycle, a method frequently used in other parts of the world. Historic reality may be much more complicated, as Koyama showed in this Japanese case. Koyama's study also demonstrates the inaccuracy of the conventional view of the traditional Japanese family system as lineally extended. In his sample, only a little more than one-third, 34.896, of the families took the lineally extended form at any one time. A larger proportion, 40.5%, assumed the nuclear form, which emphasizes the fallacy in the common belief that the emergence of the in Japan is a phenomenon solely of the industrial age. This misconception results from not distinguishing between the two functionally different types of nuclear families, namely, the cor- porate, stem family, which can take the nuclear form in the structural sense at one time or another in its developmental cycle; and, on the other hand, the , which usually but not necessarily takes the structurally nuclear form. The increased incidence in modern Japan of the nuclear JAPAN IN BEHAVIORAL SCIENCES 115 family is caused by two factors operating simultaneously: (1) the more or less "natural" increase in stem families of the nuclear form because of the population explosion, i.e., the increase in the number of junior who have had to.set up their own neolocal families; and (2) the processes of industrialization and urbanization, which have encouraged the emergence of the conjugal type of nuclear family. In Japan, the concept of family cycle has become an important, and promising, theoretical tool in the socioIogica1 study of the family. Studies have been made of attempts to gauge the relative economic well-being of a family at different stages of a cycle. They examine the changes in the roles of individual family members, who are maturing and aging at different rates, and correlate marital discords with the different stages in the family cycle (Matsubara, et a1 1967: 57-74; Morioka 1965, 1967; Suzuki 1940). 4. Kinship Roles In the past, analyses of kin roles and values were limited on the whole to accounts of normative roles in the traditional system, such as roles of the head of the , the , the , and the -in-law. More recently, however, sociologists have turned their attention to statistical surveys and obtained data on both actual kinship practices and normative preferences. Koyama's comparative survey (1960) of rural, urban, and intermediate families shows differences in the attitudes toward tradi- tional values with respect to the family system, such as the care of the , privileges and duties of the senior , and of a man or woman (yoshi)to carry on the family. (Four groups were actually sampled, since families in the intermediate category were divided into "farmers" and "non-farmers.") The results were as expected, that the rural sample is more traditional than the urban sample. More revealing are studies of division of labor among family members, the power relationship between and , relative marital satisfac- tion and its correlates, and other familial relations (Masuda 1966; Matsu- bara 1966). These studies, like many others, are directly influenced by developments in American family sociology, especially works of Parsons and Bales, Burgess and Loacke, Sussman, Wolf, Blood, and others. Blood's work (1967) is of special interest since it explicitly compares a Tokyo sample with a Detroit sample, using comparable questionnaires insofar as cultural differences would allow. The finding that couples from Tokyo did not see their relatives so often as did the Detroit couples, for example, is well worth thinking about: 75% of the Tokyo sample saw rela- tives no more than "a few times a month," whereas 71% of the Detroit sample saw them more than "once or twice a week." All kinship roles in Japan are changing-generally in the egalitarian direction-with the result that younger and female members of the family 116 RICE UNIVERSITY STUDIES now enjoy higher status than they did before the war, at the expense of the older and male members. Not surprisingly, the person taking the hardest blow in this changing time is the senior male of the family, particularly after his earning power has declined, as on retirement (Morioka 1967; Takenaka 1966). Since the moral prescription to support one's parents is no longer a compelling force, many young people now refuse to do so. But the problems of the aged are not simply economic; they are also social and psychological. Aged people find it difficult to accept the prospects that their children will no longer assist them financially or that they will no longer accord them the respect and deference the aged feel is due them. Many old people are forced to bear either the loneliness of living by them- selves or the hardship of Iiving with children who accept them on suf- ferance, not as a matter of right. The realization that they are not wanted and not needed is intolerable in a society like Japan's where affective interdependence has always been the essence of life.

5. Social Problems Social problems related to kinship are numerous. The rapid develop- ment of programs of social welfare in government agencies and the estab- lishment of social welfare curricula at colleges and universities in recent years are evidence of the troubled times in Japanese society. One set of problems, those of the aged, has been mentioned above. A few others are of note. In sharp contrast with the United States, rates in Japan before the war had generally declined since 1883, when statistical records became available, changingfrom 3.39 to 0.66 per thousand persons in 1943. Postwar years saw a sharp rise in the rate, to 1.02 in 1947, and a subsequent drop to slightly below 0.80 in the late 1950's. This level was maintained at least until 1965. Two facts are outstanding: 1) the postwar rates have never become as low as the lowest prewar rates, and 2) the rates seem to have become stabilized in the last several years. Kawashima and Steiner (1960: 239) have predicted, on the basis of data up to 1957, that "Japan will in time follow more closely a pattern which is familiar to us elsewhere." Is Japan now at the "launching stage" for a sharply rising trend in the divorce rate? Whatever the answer to this question may be, it is clear that the phenome- non of divorce is radically different from what it was before the war in such aspects as motivation, Iegal grounds, and procedures of settlement. The fact that in 1964 more than twice as many were initiated by than by is sufficient to indicate the degree of departure from the prewar days (Shikata 1966: 259). We have no idea how prevalent fatherless families were in Japan before the war and to what extent they were an object of social concern. In the JAPAN IN BEHAVIORAL SCIENCES 117

postwar years, the problem of the "incomplete family" (kesson kazoku), as fatherless or motherless families have been labeled, has risen. Fatherless families, in particular, pose problems because of the low earning power of women, in general, and the still lower earning power of who must devote a considerable portion of their time to their children. The necessity for many of them to work outside the home, and for their older children to contribute to the family income, has been seen as causing the rise of juvenile delinquency and other problems of children (Mitsukawa 1966). According to a 1956 survey, 420,000 of the 17,390,000 households in Japan consisted of a mother with children less than eighteen years of age (Takada 1958). Other data show that a large number of fatherless families are on welfare. This fact is in keeping with the lowly occupations and small incomes these families generally have. When we turn to the reasons these families become fatherless, we find that Takada's study uncovered a signifi- cant statistical comparison: desertions by husbands and unmarried mothers were so few that they were included in the miscellaneous category, which totaled 16.2%. Over 80% were made fatherless by death and divorce (67.6% and 16.2%, respectively). Western studies claim that the fatherless family, called the "," is characterized by a complex of traits that prominently include: 1) a preponderance of this phenomenon in the lower social class, and 2) an emphasis by "" on sexual exploitation without concern for a permanent or semipermanent attachment, which frequently results in desertion by the man and an unwed mother (Lewis 1966: xlii-lii; Smith 1956). Japan's fatherless families may be "matrifocal" in form and are in- deed more common in the lower class, but there is no evidence among them of frequent desertion by men, or that extra- and premarital pregnancies are common. The Japanese findings seem to require reexamination of the concept of the matrifocal family. 6. Changes in Kinship These considerations lead us to the topic of kinship change. This topic can be divided into long-range changes in the history of the Japanese family, , and other matters of kinship, and short-range changes since the Meiji Restoration, World War 11, and the past few years. Most studies of the long-range changes are of necessity based on analysis of historical documents, whereas studies of more recent changes utilize a number of empirical techniques as well. Obtaining data from the past so that legitimate comparisons can be made is sometimes a problem. Let me com- ment on a few of the methods used in these studies. Some studies (e.g., Dore 1958) see the observable pattern of the present family system as resulting from changes from past norms. They compare an 118 RICE UNIVERSITY STUDIES ideational order of the past with a phenomenal order of the present. The reliability of this procedure is questionable since the two orders of social structure may not change hand in hand, as is assumed. A second way of studying change is exemplified in Koyama's rural- urban comparison, previously mentioned. Here the rural sample is assumed to have retained the traditional family ideology to a great extent, whereas the urban sample is assumed to manifest strongly the effects of urbanization. This procedure is more satisfactory than the one previously discussed, since (rural) norms are actually compared with (urban) norms; but there is still a question of the extent to which present rural norms in fact coincide with past traditionaI norms so that change may be inferred. A third method is represented by Blood's study (1967), in which he assumes that the marital state (as judged by such matters as the degree of the husband's appreciation of the wife) of a couple married for two years at the time of the study will be the same in three years as the marital state of a couple married for five years at the time of the study. Using this assumption, Blood is able to assert (1967: 69) that "by the second decade of marriage, fewer miai husbands are still chivalrous" to their wives. But the fact may be that husbands who are married more than.a decade, being older and more old-fashioned, may have never been very chivalrous. In short, Blood assumes that insofar as marital roles are concerned the society is unchanged. When a society is known to be going through rapid change, it is dangerous to assume that ten years hence a newly married couple will be acting the same way as a couple who has already been married ten years is now acting.

7. Conclusion I wish to point out some subjects of research on Japanese kinship that have been relatively neglected and to suggest a few directions in which research might proceed. One such area is the formal semantic analysis of Japanese kinship termi- nology. Several conventional analyses of Japanese kin terms have been made (Befu and Norbeck 1958; Norbeck 1963; Smith 1962) but only one formal semantic analysis, and it is small in scope (Wallace 1965). Additional semantic analysis should bring into relief salient features of the Japanese kinship system and may tell us new and important things about the system. The subject of fatherless families in the lower class discussed above points to the need to study variations in the Japanese kinship structure along lines other than the rural-urban and regional contrasts-for example, differences by social and occupational class. Instructive here would be more information about differences in life style between families of independent businessmen and salaried employees. Some studies, notably those by Caudill, have correlated various psychological syndromes with membership JAPAN IN BEHAVIORAL SCIENCES 119

in these two types of families (Caudill and Plath 1966), but I have yet to see a serious sociological analysis of the life styles of such families. A third direction in which Japanese kinship study might proceed with profit is to relate family activities to allocation of economic resources, such as time, space, money, and other possessions (Matsubara, et a1 1967: 67-74). What rights and responsibilities do family members have in the use of these resources? How much of which resources are used separately and how much together? Although anthropologists are well aware of the importance of family economics in agrarian and primitive societies, they tend to neglect this subject when dealing with modern urban society, relegating it to the domain of economics qua economics. I think we would better under- stand the functioning of the kinship system if we studied it as an economic system. I shall close with a few remarks of a broader nature. Reviewing the literature on Japanese kinship and the family, I am struck by the lack of cooperation between Japanese anthropologists and Japanese sociologists. There is reluctance on the part of anthropologists to study urban families, for example, and little effort is made by sociologists to look into rural practices of kinship. They avoid invading each other's "territory" for em- pirical research, and neither side seems to take advantage of the findings of the other in order to integrate knowledge about the Japanese family and kinship. A closer relationship between the two groups should make signifi- cant contributions to knowledge. Two exceptional and important studies by Ariga (1939) and Nakano (1964), sociologists who "did ," demonstrate the fruitfulness of such a close relationship. In attempting to understand kinship in Japan today, we are not helped by trying to see it as a disintegrated traditional system or as an incom- pletely Westernized system. Japanese kinship today is a unique system with dynamics of its own, resulting from a heterogeneous tradition and a host of modern factors, such as industrialization, urbanization, and dif- fusion of Western ideology. For years to come, it will remain an amalgam of these forces and will have to be understood as such. Facile comparison with the West, and the ready assumption that Japanese kinship is changing in the Western direction, will not increase our understanding. It may be added that what is true of kinship analysis is also true of Japanese studies as a whole. RICE UNIVERSITY STUDIES

REFERENCES ARIGA, KIZAEMON 1939 Nambu Ninoe-gun Ishigama-mura ni okeru daikazoku seido to nago seido. Tokyo, Attlc Museum Report 43. 1947 Dozoku to shlnzoku- In Shinobu Origushl, ed. N~honminzokugaku no tame nl 2. Tokyo, Minkan Densho no Kai. Pp. 1-70. BEFU, HARUMI 1963 Patrilineal descent and personal kindred in Japan. American Anthropologist 65: 1328-1341. BEFU, HARUMI, and EDWARD NORBECK 1958 Japanese usages of terms of relationship. Southwestern Journal of Anthropology 14: 66-88. BLOOD, ROBERT O., JR. 1967 match and arranged marriage, a Tokyo-Detroit comparison. New York, Free Press. BROWN, KEITH 1966 Dozoku and the ldeology of descent in rural Japan. American Anthropologist 68: 1129-1151, 1968 The content of dozoku relationships in Japan. Ethnology 7: 113-138. CAUDILL, WILLIAM, and DAVID W. PLATH 1966 Who sleeps by whom? - involvement in urban Japanese families. Psy- chiatry 29: 344-366. DORE, RONALD P. 1958 City life in Japan. Berkeley and Los Angeles, University of California Press. GAMO, MASAO 1962 Shinzoku. In Nihon minzokugaku talkel 3.Toky0, Helbonsha. Pp. 233-258. K.AWASHIMA, TAKEYOSHI, and KURT STEINER 1960 Modern~zationand divorce rate trends In Japan. In Thomas C. Smith, ed. City and village in Japan (Economic Development and Cultural Change 9: No. 1, Pt. 2). Univers~tyof Chicago Press. Pp. 213-239. KITANO, SElICHI 1966 Nlhon noson ni okeru dozoku to shinzoku. In Dai kyukai kokusai kazoku kenkyu semlna hokokusho. Tokyo, Nihon Yunesuko Kokunai Iinkal. Pp. 36-41. KOYAMA, TAKASHI 1959 Kazuko keltai no shuki-teki henka. In Seiichi Kitano and Yuzuru Okada, eds. Ie- sono kozo bunseki. Tokyo, Sobunsha Pp. 67-83. KOYAMA. TAKASHI, ed. 1960 Gendai kazoku no kenkyu. Tokyo, Kobundo. LEWIS, OSCAR 1966 La vida. New York, Random House. MASUDA, KOKICHI 1966 Kazoku no na~bukozo. In Kaoru Ohashi and Kokichi Masuda, eds. Kazoku shakaigaku. Tokyo, Kawashima Shoten. Pp. 26-49. MATSUBARA, HARUO 1966 Kazoku-na~no ningen kankei. In Kaoru Ohashi and Kokichi Masuda, eds. Kazoku shakalgaku. Tokyo, Kawashima Shoten. Pp. 69-94. MATSUBARA, HARUO, et al 1967 Gendai Nihon no shakaigaku. Tokyo, Jichosha MITSUKAWA, HARUYUKI 1966 Kesson kazoku. In Kaoru Ohashi and Kokichi Masuda, eds. Kazoku shakaigaku. Tokyo, Kawashlma Shoten. Pp. 210-234. JAPAN IN BEHAVIORAL SCIENCES

MORIOKA, KIYOMI 1965 Kazoku no hendo ni taio seru shuki dankai no settei. Shakai Kagaku Janaru 6: 3 17-349. 1967a Life cycle patterns in Japan, China, and the Untted States. Journal of Marriage and the Family 29: 595-606. 1967b The social network around old people In a suburb of Tokyo. Internattonal Christian University. The Ftnal Report of the Ford Research Project. Pp. 237-252. NAKANE, CHIE 1967 Kinship and economic organization in rural Japan. London, Athlone. NAKANO, TAKASHI 1957 Kazoku to shinzoku, In Tadashi Fukutake, ed. Koza shakaigaku 4. Tokyo, Tokyo University Press. Pp. 44-70. 1964 Shoka dozokudan no kenkyu: noren o meguru ie no kenkyu. Tokyo, Miraisha. NORBECK, EDWARD 1963 Lewis Henry Morgan and Japanese terms of relationship: profit through error. Southwestern Journal of Anthropology 19: 208-215. OIKAWA, HIROSHI 1939 Dozoku soshtki to kon'in oyobt soso no girei. Minzokugaku Nempo 2: 1-40. SHIKATA, HISAO 1966 Rikon iede. In Kaoru Ohasht and Koktchi Masuda, eds. Kazoku shakaigaku. Tokyo, Kawashima Shoten. Pp. 256-279. SMITH, RAYMOND T. 1956 The Negro family in Bntish Guiana: family structure and social status in the villages. London, Routledge and Kegan Paul. SMITH, ROBERT J. 1962 Japanese kinship term~nology:the history of a nomenclature. Ethnology 1: 349-359. SUZUKI, EITARO 1940 Nihon noson shakatgaku genri. Tokyo, Nihon Hyoronsha. 1942 Nihonjtn kazoku no sedai-teki hatten ni okeru shuki-teki ritsudosei ni tsuite. In Teizo Toda and Eitaro Suzuki, eds. Kazoku to sonraku 2. Tokyo, Nikko Shoin. Pp. 1-50. TAKADA, HIROMICHI 1958 Hahako setai to sono taisaku. In Hyoe Ouchi, ed. Roreisha hahako no jittai. Tokyo, Toyo Ketzai Shtmposha. Pp. 156-18 1. TAKENAKA, KAZUO 1966 Ronensha mondai. In Kaoru Ohashi and Kokichi Masuda, eds. Kazoku shakaigaku. Tokyo, Kawashima Shoten. Pp. 235-255. WALLACE, ANTHONY F. C. 1965 The problem of the psychological validtty of componential analysis. In Eugene A. Hammel, ed. Formal semantic analysis. Wash~ngton,D.C. American Anthropological As~ociation. Pp, 229-248. (Originally published in American Anthropologtst 67: 5, pt 2.)